CAREER: Sustaining Scientific Infrastructure: Researching Transition from Grants to Peer Production

University Of Texas At Austin, Austin TX

Investigators

Abstract

Software is becoming increasingly critical to the activities of science, yet securing long term support for this infrastructure is becoming more challenging. Given constrained budgets, the options for effectively and efficiently supporting sustainable software infrastructure are of growing importance. This project examines the transition of research infrastructure from grant-based funding to long term sustainable models, by examining scientific software projects as they transition to an open source peer production model. The project tracks software packages as they attempt to transition, and develops a theoretical framework for understanding this process and for explaining their success or failure. The outcome of this research will inform science policy by shaping how the community funds software infrastructure projects in their initial phases, and how to encourage the transition to long-term models thereafter. Inspired by the success of open source software, funding agencies are encouraging transitions to forms of organization that organizational researchers call 'peer production.' Yet peer production research has shown that simply making code available under an open source license is insufficient to build a motivated and productive community. Successful transition includes changes in team structure (from local to distributed), collaboration technologies (from controlled to open), governance (from hierarchical to shared), and management of participants (from predictable to unpredictable commitment). This project examines successful and unsuccessful transitions from initial grant funding to peer production in scientific software development. Detailed case study histories are developed, by interviewing project participants, funders, and users of scientific software, and by analyzing project archives and software repositories. The research includes six transitioning projects: yt, Enzo, Eclipse PTP, IRODs, Apache OODT and Airavata. In addition, a panel of projects funded by the NSF Software Infrastructure for Sustained Innovation program (which requires plans for sustainability) will be tracked over time to study successful transitions, unsuccessful attempts, and cases where transition was judged inappropriate. The research is combined with an integrated education plan designed to build organizational skills in the scientific community to enable successful transitions to peer production.